A low-complexity energy disaggregation method: performance and robustness

Hana Altrabalsi, Jing Liao, Lina Stankovic, Vladimir Stankovic

Research output: Chapter in Book/Report/Conference proceedingConference contribution book

3 Citations (Scopus)

Abstract

Disaggregating total household's energy data down to individual appliances via non-intrusive appliance load monitoring (NALM) has generated renewed interest with ongoing or planned large-scale smart meter deployments worldwide. Of special interest are NALM algorithms that are of low complexity and operate in near real time, supporting emerging applications such as in-home displays, remote appliance scheduling and home automation, and use low sampling rates data from commercial smart meters. NALM methods, based on Hidden Markov Model (HMM) and its variations, have become the state of the art due to their high performance, but suffer from high computational cost. In this paper, we develop an alternative approach based on support vector machine (SVM) and k-means, where k-means is used to reduce the SVM training set size by identifying only the representative subset of the original dataset for the SVM training. The resulting scheme outperforms individual k-means and SVM classifiers and shows competitive performance to the state-of-the-art HMM-based NALM method with up to 45 times lower execution time (including training and testing).
LanguageEnglish
Title of host publication2014 IEEE Symposium on Computational Intelligence Applications in Smart Grid (CIASG)
Place of PublicationPiscataway, NJ.
PublisherIEEE
Pages1-8
Number of pages8
DOIs
Publication statusPublished - 9 Dec 2014
EventIEEE Symposium on Computational Intelligence Applications in Smart Grid - Orlando, United States
Duration: 9 Dec 201412 Dec 2014

Conference

ConferenceIEEE Symposium on Computational Intelligence Applications in Smart Grid
Abbreviated titleCIASG 2014
CountryUnited States
CityOrlando
Period9/12/1412/12/14

Fingerprint

Support vector machines
Smart meters
Monitoring
Hidden Markov models
Classifiers
Automation
Display devices
Scheduling
Sampling
Testing
Costs

Keywords

  • complexity theory
  • feature extraction
  • hidden Markov models
  • home appliances
  • support vector machines

Cite this

Altrabalsi, H., Liao, J., Stankovic, L., & Stankovic, V. (2014). A low-complexity energy disaggregation method: performance and robustness. In 2014 IEEE Symposium on Computational Intelligence Applications in Smart Grid (CIASG) (pp. 1-8). Piscataway, NJ.: IEEE. https://doi.org/10.1109/CIASG.2014.7011569
Altrabalsi, Hana ; Liao, Jing ; Stankovic, Lina ; Stankovic, Vladimir. / A low-complexity energy disaggregation method : performance and robustness. 2014 IEEE Symposium on Computational Intelligence Applications in Smart Grid (CIASG). Piscataway, NJ. : IEEE, 2014. pp. 1-8
@inproceedings{3cc8c0e377e24e54afd198bdb7cc4c55,
title = "A low-complexity energy disaggregation method: performance and robustness",
abstract = "Disaggregating total household's energy data down to individual appliances via non-intrusive appliance load monitoring (NALM) has generated renewed interest with ongoing or planned large-scale smart meter deployments worldwide. Of special interest are NALM algorithms that are of low complexity and operate in near real time, supporting emerging applications such as in-home displays, remote appliance scheduling and home automation, and use low sampling rates data from commercial smart meters. NALM methods, based on Hidden Markov Model (HMM) and its variations, have become the state of the art due to their high performance, but suffer from high computational cost. In this paper, we develop an alternative approach based on support vector machine (SVM) and k-means, where k-means is used to reduce the SVM training set size by identifying only the representative subset of the original dataset for the SVM training. The resulting scheme outperforms individual k-means and SVM classifiers and shows competitive performance to the state-of-the-art HMM-based NALM method with up to 45 times lower execution time (including training and testing).",
keywords = "complexity theory, feature extraction, hidden Markov models, home appliances, support vector machines",
author = "Hana Altrabalsi and Jing Liao and Lina Stankovic and Vladimir Stankovic",
note = "(c) 2014 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other users, including reprinting/ republishing this material for advertising or promotional purposes, creating new collective works for resale or redistribution to servers or lists, or reuse of any copyrighted components of this work in other works.",
year = "2014",
month = "12",
day = "9",
doi = "10.1109/CIASG.2014.7011569",
language = "English",
pages = "1--8",
booktitle = "2014 IEEE Symposium on Computational Intelligence Applications in Smart Grid (CIASG)",
publisher = "IEEE",

}

Altrabalsi, H, Liao, J, Stankovic, L & Stankovic, V 2014, A low-complexity energy disaggregation method: performance and robustness. in 2014 IEEE Symposium on Computational Intelligence Applications in Smart Grid (CIASG). IEEE, Piscataway, NJ., pp. 1-8, IEEE Symposium on Computational Intelligence Applications in Smart Grid, Orlando, United States, 9/12/14. https://doi.org/10.1109/CIASG.2014.7011569

A low-complexity energy disaggregation method : performance and robustness. / Altrabalsi, Hana; Liao, Jing; Stankovic, Lina; Stankovic, Vladimir.

2014 IEEE Symposium on Computational Intelligence Applications in Smart Grid (CIASG). Piscataway, NJ. : IEEE, 2014. p. 1-8.

Research output: Chapter in Book/Report/Conference proceedingConference contribution book

TY - GEN

T1 - A low-complexity energy disaggregation method

T2 - performance and robustness

AU - Altrabalsi, Hana

AU - Liao, Jing

AU - Stankovic, Lina

AU - Stankovic, Vladimir

N1 - (c) 2014 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other users, including reprinting/ republishing this material for advertising or promotional purposes, creating new collective works for resale or redistribution to servers or lists, or reuse of any copyrighted components of this work in other works.

PY - 2014/12/9

Y1 - 2014/12/9

N2 - Disaggregating total household's energy data down to individual appliances via non-intrusive appliance load monitoring (NALM) has generated renewed interest with ongoing or planned large-scale smart meter deployments worldwide. Of special interest are NALM algorithms that are of low complexity and operate in near real time, supporting emerging applications such as in-home displays, remote appliance scheduling and home automation, and use low sampling rates data from commercial smart meters. NALM methods, based on Hidden Markov Model (HMM) and its variations, have become the state of the art due to their high performance, but suffer from high computational cost. In this paper, we develop an alternative approach based on support vector machine (SVM) and k-means, where k-means is used to reduce the SVM training set size by identifying only the representative subset of the original dataset for the SVM training. The resulting scheme outperforms individual k-means and SVM classifiers and shows competitive performance to the state-of-the-art HMM-based NALM method with up to 45 times lower execution time (including training and testing).

AB - Disaggregating total household's energy data down to individual appliances via non-intrusive appliance load monitoring (NALM) has generated renewed interest with ongoing or planned large-scale smart meter deployments worldwide. Of special interest are NALM algorithms that are of low complexity and operate in near real time, supporting emerging applications such as in-home displays, remote appliance scheduling and home automation, and use low sampling rates data from commercial smart meters. NALM methods, based on Hidden Markov Model (HMM) and its variations, have become the state of the art due to their high performance, but suffer from high computational cost. In this paper, we develop an alternative approach based on support vector machine (SVM) and k-means, where k-means is used to reduce the SVM training set size by identifying only the representative subset of the original dataset for the SVM training. The resulting scheme outperforms individual k-means and SVM classifiers and shows competitive performance to the state-of-the-art HMM-based NALM method with up to 45 times lower execution time (including training and testing).

KW - complexity theory

KW - feature extraction

KW - hidden Markov models

KW - home appliances

KW - support vector machines

UR - http://www.ieee-ssci.org/CIASG.html

UR - http://ciasg.rtpis.org/

U2 - 10.1109/CIASG.2014.7011569

DO - 10.1109/CIASG.2014.7011569

M3 - Conference contribution book

SP - 1

EP - 8

BT - 2014 IEEE Symposium on Computational Intelligence Applications in Smart Grid (CIASG)

PB - IEEE

CY - Piscataway, NJ.

ER -

Altrabalsi H, Liao J, Stankovic L, Stankovic V. A low-complexity energy disaggregation method: performance and robustness. In 2014 IEEE Symposium on Computational Intelligence Applications in Smart Grid (CIASG). Piscataway, NJ.: IEEE. 2014. p. 1-8 https://doi.org/10.1109/CIASG.2014.7011569